Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

Identification of key regulatory pathways and regulators in the

pathogenesis of hepatitis C virus-induced hepatocellular carcinoma

Guolin Chen1, Yongguo Li1, Yingjie Ma1, and Xue Tan1

1First Affiliated Hospital of Harbin Medical University

May 5, 2020

Abstract

Hepatitis virus infection is a leading cause of chronic liver diseases, including cirrhosis and hepatocellular carcinoma (HCC).

However, the molecular mechanism by which hepatitis causes liver cancer remains unclear. Additionally, new biomarkers for

diagnosis, prognosis and therapeutics are needed. Regulatory pathways play important roles in many pathogenic processes,

and identifying the pathways by which hepatitis C virus (HCV) induces HCC may lead to better diagnosis and treatment. We

employed a systematic approach to identify important regulatory pathways in this disease process, and found several important

regulators. First, three networks were constructed based on the gene expression in patients with hepatitis alone, HCC alone,

and hepatitis with HCC. A priority algorithm was used to extract the regulatory pathways from the networks, which were then

scored based on the disease-related genetic information to identify key pathways. After integrating the regulatory pathways

involved in the three networks, we found key regulatory genes, including EZH2 and hsa-miR-155-5p. Based on network analysis,

it appeared that in HCC patients the abnormal expression of genes and miRNAs were mostly caused by abnormal expression

of these key regulatory factors. This method may help researchers discover the potential pathogenic factors of HCC and could

also yield new biomarkers for disease diagnosis.

Introduction

Hepatocellular carcinoma (HCC) is the second leading cause of cancer-related deaths worldwide 1. Thedisease is often diagnosed at an advanced stage and progresses rapidly. Therefore, early diagnosis is veryimportant to improve the prognosis of patients. Currently, early clinical screening methods involve serumalpha fetoprotein (AFP) detection and liver ultrasound examination2. However, the sensitivity and specificityof markers such as AFP are marginal 3. Additionally, ultrasound examination relies heavily on the subjectivejudgment of the operator, and conventional ultrasound often does not give results which can be used toconclusively identify liver lesions. Therefore, there is an urgent need to find more effective and accuratemethods for screening for liver cancer. As the understanding of tumor biology improves, liquid biopsy willbecome an increasingly useful tool for early diagnosis. Risk factors for primary liver cancer include hepatitisB virus (HBV) or hepatitis C virus (HCV) infection, aflatoxin B intake, alcohol consumption, cirrhosis andothers. Among these risk factors, HBV and HCV infections are the most significant, with viral hepatitis(HBV and HCV infections) accounting for 90% and 40% of liver cancer incidence in developing and developedcountries, respectively.

To date, few studies have been conducted to assess the factors leading to liver cancer in HCV patients. Atpresent, HCV RNA, cirrhosis and HCV genotype are thought to affect the occurrence of HCV-related livercancer, but these factors have not been conclusively proven. At present, 180 million people are chronicallyinfected with HCV, which has been reported to cause more than 350,000 deaths annually 4. Epidemiologicalstudies also indicate that HCV is associated with a number of extrahepatic manifestations including insulinresistance, type 2 diabetes mellitus, glomerulopathies, oral manifestations and others5-7.

1

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

It has been demonstrated that 55% to 85% of new HCV-infected patients will develop chronic hepatitis C,and 20% to 30% of chronically ill patients will develop cirrhosis and liver failure 8. Over the course of 30years, 1% to 3% of patients with HCV without cirrhosis will eventually develop HCC 9,10. Studies haveshown that one-third of HCC cases are caused by hepatitis C11. At present, there are three major knownmechanisms for HCV-induced HCC: direct pathways involving HCV core proteins, indirect pathways causedby oxidative stress and steatosis, and microRNA (miRNA)-related pathways 12. Previous studies have shownthat while biological signaling systems are complex, the analysis of linear pathways can still provide valuableinsights13.

In this study, we integrated multiple resources, including KEGG pathways, known disease genes, miRNAsand differentially expressed genes, to identify regulatory pathways and key regulators in HCV and HCC fromcurated trans factor and miRNA regulatory networks. This analysis revealed that EZH2 and hsa-miR-155-5p are critical genes in the development of hepatocellular carcinoma. These genes and pathways may beimportant biomarkers for predicting HCV-induced HCC. The workflow of our study is shown in Fig. 1.

Materials and Methods

Construction of human Transcription factor and miRNA regulatory networks

The human Transcription factor (TF) and miRNA regulatory networks were built by integrating miRTarBase,TarBase, TRANSFAC and TransmiR14-16. These four databases include curated interactions among humanTFs, miRNAs, and target genes as well as standardization of gene and miRNA names within the regulatorynetworks using data from NCBI and miRbase databases. Additionally, all regulatory relationships withinthe regulatory network were literature-supported. In total, there were 460 TFs, 2,434 miRNAs, 13,898 targetgenes and 98,894 edges in the regulatory network.

Known HCC and HCV-associated genes and miRNAs

DisGeNET, a discovery platform containing one of the largest publicly available collections of genes andvariants associated with human diseases, was utilized to identify two disease-associated genes 17. Two disease-associated miRNAs were collected from the miR2Disease18 and HMDD 19 , which are curated databasescontaining experimental evidence for human microRNA (miRNA) and disease associations. We also utilizedgenes in the KEGG pathways associated with HCC (168) or HCV (155). We included 30 known HCC-associated genes in DisGeNET and 463 known HCC-associated miRNAs from either miR2Disease or HMDD.Finally, 18 known HCV-associated genes in DisGeNET and 100 known HCV-associated miRNAs in eithermiR2Disease or HMDD were used for network analysis.

Disease-related network construction

For the disease-related network construction, the closer the nodes in the network to the known disease genes,the more likely they are disease-associated 20. In order to construct a more closely related subnet, we selectednodes directly connected to the known disease-associated genes in the background network to build an HCCand HCV-related network. In total, there were 409 TFs, 2,300 miRNAs, 10,697 target gene and 48423 edgesin this regulatory network.

Differentially expressed genes in the three datasets

The normalized mRNA expression profiles of HCC (TCGA), HCV (GSE15387) and HCV-related HCC(GSE44074) were downloaded from the Gene Expression Omnibus (GEO) database 21 and The CancerGenome Atlas (TCGA) database 22. There were 374 HCC samples and 50 normal samples in the TCGAdata set, 35 HCV-related HCC samples and 37 HCC samples in the GSE44074, as well as 60 HCV samplesand 60 normal samples in the GSE15387. For mRNA expression data, probe sets were mapped to EntrezGene IDs. When multiple probes corresponded to the same gene, the mean expression value of these probeswas used to represent the gene expression level. We obtained 2, 3, and 4 differentially expressed genes atthe p -values of less than 0.05 by using edgeR (TCGA data) and SAM (GEO data) in each of the three datasets.

2

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

Identification of the subnetworks for each dataset

To construct subnetworks for each dataset, we extracted differentially expressed genes and their neighborgenes from the disease-related network. The regulatory relationships between these genes and miRNAs con-stituted a core regulatory subnetwork at multiple stages of disease development. We identified 3 subnetworks,which we termed the HCC subnetwork, HCC-HCV subnetwork and HCV subnetwork.

Extraction of candidate risk regulatory pathways

Using the BFS algorithm to extract risk regulatory pathways from the three subnetworks, we identified allthe pathways in the network from the nodes indegree 0 to outdegree 0, and pathways with a length greaterthan 2 were regarded as the candidate risk pathways.

Prediction of key regulators

Gene expression varies in different tissues and during different diseases. Some genes are expressed at a specificstage of a given disease, while some genes continue to play a role throughout the process. We analyzed allthe pathways in the three subnetworks to identify the most critical pathways in each network by examininghighly shared genes. We propose a KP score to evaluate key pathways, which is calculated as follows:

Where denotes the number of nodes on a pathway within a subnetwork, denotes the number of intersectionnodes between pathway and subnetwork , denotes the length of the longest pathway within the subnetworkthat satisfies the conditions, denotes location weight score of the intersection gene within the pathway,denotes whether the gene at this position is an intersection gene, if yes, then the value of is 1, if not, thevalue of is 0, upstream genes get higher scores.

Survival analysis

In this study, we constructed three subnetworks for HCC, HCV samples and normal samples, and identifiedkey pathways from the subnetworks. We next investigated whether the key regulators could distinguish HCCpatients with good or poor outcomes. From these data, we obtained TCGA HCC dataset mRNA expression,miRNA expression and clinical information. Next, we used the K-means method (K=2) to cluster all patientsinto two groups based on the mRNA and miRNA expression. Finally, Kaplan–Meier curve and log-rank testswere used to evaluate the difference in overall survival time between the two groups of patients.

Results and Discussion

Three diseases-related subnetworks

We used three differentially expressed genes to construct three disaease-related subnetworks based on thedisease background network, which contained 409 TFs, 2300 miRNAs, 10,697 target gene and 48,423 edges.We then mapped the 247, 237, and 103 differentially expressed genes obtained from the three disease-related differential expression datasets to the background network to obtain three subnetworks. The HCC-related TF-miRNA regulatory subnetwork included 228 edges and 169 nodes, including 22 TFs, 95 miRNAs,and 52 target genes. The HCC-related TF-miRNA regulatory subnetwork included 911 edges and 464nodes including 46 TFs, 236 miRNAs, and 182 target genes. The HCC-HCV-related TF-miRNA regulatorysubnetwork included 513 edges and 307 nodes, including 29 TFs, 157 miRNAs, and 121 target genes (Fig.S1).

Gene ontology (GO) and KEGG functional enrichment analyses were performed to identify the significantlyenriched biological processes and pathways in the three subnetworks using DAVID 23online tools to performenrichment analysis. The significantly enriched results (FDR

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

signaling pathway. Many of these pathways have previously been implicated in HCC and HCV diseases24-27.

The risk regulatory pathways and key regulators

The BFS method was used to traverse three subnetworks to obtain all pathways in the network with anin-degree of 0 to out-degree of 0, with pathway lengths required to be longer than two nodes. The HCCsubnetwork obtained 5,284,069 pathways, the HCC-HCV subnetwork contained 929 pathways, and the HCVsubnetwork contained 235 pathways. Each pathway contained several subnetworks, and elucidating keyregulators that play important roles in the development of the disease requires identifying the most importantof these. To this end, we used KP scores to screen the 5 highest scoring pathways in all the subnetworks(Table 1). All the regulatory pathways were integrated to obtain a network of HCV and HCC processes.From this network, we found that HCV-related genes were mainly enriched in the upstream nodes of thenetwork (green background in Fig. 3), while the genes affecting HCC and HCV were mainly enriched in themiddle regions of the network (yellow background in Fig. 3). Finally, genes related to HCC appeared in thedownstream regions of the network (violet background in Fig. 3). The network structure also reflected therole of inflammation in carcinogenesis, as many genes associated with inflammatory factors linked nodes inthe HCC and HCV networks. To test whether abnormal expression of some core genes at the center of thenetwork affected patient outcome, we examined hsa-miR-155-5p , FOXM1 ,EZH2 more closely.

EZH2 and hsa-miR-155-5p are key regulators

We further analyzed the core genes in the network, and found thatFOXM1 , EZH2 , E2F1 and hsa-miR-93-5p were significantly correlated with the occurrence of HCC in HCV patients (Fig. S2). Out of these genes,EZH2 was the most downstream and directly regulated the hub node for hsa-miR-155-5p within the network(Fig. S3). This indicated that EZH2 may be an important gene implicated in the transition from HCV toHCC. Other research has found that EZH2 and hsa-miR-155-5p may play important roles in the progressionof both HCC and HCV 28-31, which fits with works indicating that hsa-miR-155-5p participates in HCV-induced HCC processes 32,33. High expression ofEZH2 and hsa-miR-155-5p has also been shown to correlatewith the severity of HCC 34,35. Our network analysis indicated that hsa-miR-155-5p not only plays a keyregulatory role in HCC, but also plays a role in hepatitis-induced liver cancer. Research investigating theeffects of controlling the expression ofEZH2 and hsa-miR-155-5p in HCV patients may yield new treatmentoptions.

In order to study the correlation of EZH2 andhsa-miR-155-5p expression with HCC survival rate, we com-pared the expression of these two genes in normal samples versus disease samples. Interestingly, we foundthat both EZH2 and hsa-miR-155-5pwere expressed significantly higher in tumor samples compared withnormal tissue. Between the two, we found that EZH2 was strongly correlated with the survival of patients(Fig. 4), indicating that it may be important for early diagnosis and risk prediction.

Conclusions

It is currently thought that a number of different factors influence HCV-induced HCC. However, due to thelack of appropriate models or data, it is difficult to determine the specific role of HCV in the malignanttransformation of liver cells. In order to identify and characterize these mechanisms, researchers have con-ducted genomic, transcriptomic and epigenomic studies. These studies have revealed gene mutations andgene expression changes that play a role in the development of hepatitis-induced liver cancer. Genomicresearch has found that long-term hepatitis virus infection causes significant damage, even after eradicationof hepatitis virus. Our transcriptomic data indicated that abnormal expression of certain genes and miRNAsis predictive of which patients will later develop HCC. These genes may represent biomarkers which couldenable significantly earlier detection of HCC. HCC is not only caused by hepatitis 36. Therefore, more studiesare needed to determine whether genes correlated with HCV-induced cancer are also correlated with livercancer caused by other factors.

CONFLICTS OF INTEREST

4

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

There are no conflicts to declare.

REFERENCES

1. Merte B. [The first general mathematical description of optical lenses without aberrations]. KlinischeMonatsblatter fur Augenheilkunde 1989;194(1):59-61.

2. Sato Y, Nakata K, Kato Y, Shima M, Ishii N, Koji T, Taketa K, Endo Y, Nagataki S. Early recognitionof hepatocellular carcinoma based on altered profiles of alpha-fetoprotein. N Engl J Med 1993;328(25):1802-1806.

3. Bruix J, Reig M, Sherman M. Evidence-Based Diagnosis, Staging, and Treatment of Patients WithHepatocellular Carcinoma. Gastroenterology 2016;150(4):835-853.

4. Li HC, Lo SY. Hepatitis C virus: Virology, diagnosis and treatment. World J Hepatol 2015;7(10):1377-1389.

5. Carrozzo M, Scally K. Oral manifestations of hepatitis C virus infection. World J Gastroenterol2014;20(24):7534-7543.

6. Ozkok A, Yildiz A. Hepatitis C virus associated glomerulopathies. World J Gastroenterol2014;20(24):7544-7554.

7. Montenegro L, De Michina A, Misciagna G, Guerra V, Di Leo A. Virus C hepatitis and type 2 diabetes:a cohort study in southern Italy. Am J Gastroenterol 2013;108(7):1108-1111.

8. Mahale P, Torres HA, Kramer JR, Hwang LY, Li R, Brown EL, Engels EA. Hepatitis C virus infection andthe risk of cancer among elderly US adults: A registry-based case-control study. Cancer 2017;123(7):1202-1211.

9. Huang YT, Jen CL, Yang HI, Lee MH, Su J, Lu SN, Iloeje UH, Chen CJ. Lifetime risk and sex differenceof hepatocellular carcinoma among patients with chronic hepatitis B and C. J Clin Oncol 2011;29(27):3643-3650.

10. El-Serag HB. Epidemiology of viral hepatitis and hepatocellular carcinoma. Gastroenterology2012;142(6):1264-1273 e1261.

11. Parkin DM. The global health burden of infection-associated cancers in the year 2002. Int J Cancer2006;118(12):3030-3044.

12. Tholey DM, Ahn J. Impact of Hepatitis C Virus Infection on Hepatocellular Carcinoma. GastroenterolClin North Am 2015;44(4):761-773.

13. Weng G, Bhalla US, Iyengar R. Complexity in biological signaling systems. Science 1999;284(5411):92-96.

14. Chou CH, Shrestha S, Yang CD, Chang NW, Lin YL, Liao KW, Huang WC, Sun TH, Tu SJ, Lee WH,Chiew MY, Tai CS, Wei TY, Tsai TR, Huang HT, Wang CY, Wu HY, Ho SY, Chen PR, Chuang CH,Hsieh PJ, Wu YS, Chen WL, Li MJ, Wu YC, Huang XY, Ng FL, Buddhakosai W, Huang PC, Lan KC,Huang CY, Weng SL, Cheng YN, Liang C, Hsu WL, Huang HD. miRTarBase update 2018: a resource forexperimentally validated microRNA-target interactions. Nucleic Acids Res 2018;46(D1):D296-D302.

15. Vlachos IS, Paraskevopoulou MD, Karagkouni D, Georgakilas G, Vergoulis T, Kanellos I, AnastasopoulosIL, Maniou S, Karathanou K, Kalfakakou D, Fevgas A, Dalamagas T, Hatzigeorgiou AG. DIANA-TarBasev7.0: indexing more than half a million experimentally supported miRNA:mRNA interactions. Nucleic AcidsRes 2015;43(Database issue):D153-159.

16. Tong Z, Cui Q, Wang J, Zhou Y. TransmiR v2.0: an updated transcription factor-microRNA regulationdatabase. Nucleic Acids Res 2019;47(D1):D253-D258.

5

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

17. Pinero J, Queralt-Rosinach N, Bravo A, Deu-Pons J, Bauer-Mehren A, Baron M, Sanz F, Furlong LI.DisGeNET: a discovery platform for the dynamical exploration of human diseases and their genes. Database(Oxford) 2015;2015:bav028.

18. Jiang Q, Wang Y, Hao Y, Juan L, Teng M, Zhang X, Li M, Wang G, Liu Y. miR2Disease: a manu-ally curated database for microRNA deregulation in human disease. Nucleic Acids Res 2009;37(Databaseissue):D98-104.

19. Huang Z, Shi J, Gao Y, Cui C, Zhang S, Li J, Zhou Y, Cui Q. HMDD v3.0: a database for experimentallysupported human microRNA-disease associations. Nucleic Acids Res 2019;47(D1):D1013-D1017.

20. Nibbe RK, Koyuturk M, Chance MR. An integrative -omics approach to identify functional sub-networksin human colorectal cancer. PLoS Comput Biol 2010;6(1):e1000639.

21. Barrett T, Troup DB, Wilhite SE, Ledoux P, Rudnev D, Evangelista C, Kim IF, Soboleva A, Toma-shevsky M, Marshall KA, Phillippy KH, Sherman PM, Muertter RN, Edgar R. NCBI GEO: archive forhigh-throughput functional genomic data. Nucleic Acids Res 2009;37(Database issue):D885-890.

22. Wang Z, Jensen MA, Zenklusen JC. A Practical Guide to The Cancer Genome Atlas (TCGA). MethodsMol Biol 2016;1418:111-141.

23. Huang da W, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists usingDAVID bioinformatics resources. Nat Protoc 2009;4(1):44-57.

24. Capurro MI, Xiang YY, Lobe C, Filmus J. Glypican-3 promotes the growth of hepatocellular carcinomaby stimulating canonical Wnt signaling. Cancer Res 2005;65(14):6245-6254.

25. Schmidt CM, McKillop IH, Cahill PA, Sitzmann JV. Increased MAPK expression and activity in primaryhuman hepatocellular carcinoma. Biochem Biophys Res Commun 1997;236(1):54-58.

26. Yu MW, Gladek-Yarborough A, Chiamprasert S, Santella RM, Liaw YF, Chen CJ. Cytochrome P4502E1 and glutathione S-transferase M1 polymorphisms and susceptibility to hepatocellular carcinoma. Gas-troenterology 1995;109(4):1266-1273.

27. Wu H, Ng R, Chen X, Steer CJ, Song G. MicroRNA-21 is a potential link between non-alcoholicfatty liver disease and hepatocellular carcinoma via modulation of the HBP1-p53-Srebp1c pathway. Gut2016;65(11):1850-1860.

28. Cheng YQ, Ren JP, Zhao J, Wang JM, Zhou Y, Li GY, Moorman JP, Yao ZQ. MicroRNA-155 regulatesinterferon-gamma production in natural killer cells via Tim-3 signalling in chronic hepatitis C virus infection.Immunology 2015;145(4):485-497.

29. Riad SE, El-Ekiaby N, Mekky RY, Ahmed R, El Din MA, El-Sayed M, Abouelkhair MM, Salah A,Zekri AR, Esmat G, Abdelaziz AI. Expression signature of microRNA-155 in hepatitis C virus genotype 4infection. Biomed Rep 2015;3(1):93-97.

30. Li S, Sun S, Yu S, Dong Z, Jiang K, Xiang L, Li H. Association between EZH2 Genetic Variants andHepatocellular Carcinoma in a Chinese Han Population. Clin Lab 2018;64(1):85-91.

31. Liu H, Liu Y, Liu W, Zhang W, Xu J. EZH2-mediated loss of miR-622 determines CXCR4 activation inhepatocellular carcinoma. Nat Commun 2015;6:8494.

32. Khairy A, Hamza I, Shaker O, Yosry A. Serum miRNA Panel in Egyptian Patients with ChronicHepatitis C Related Hepatocellular Carcinoma. Asian Pac J Cancer Prev 2016;17(5):2699-2703.

33. El Tayebi HM, Waly AA, Assal RA, Hosny KA, Esmat G, Abdelaziz AI. Transcriptional activation ofthe IGF-II/IGF-1R axis and inhibition of IGFBP-3 by miR-155 in hepatocellular carcinoma. Oncol Lett2015;10(5):3206-3212.

6

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

34. Fu X, Wen H, Jing L, Yang Y, Wang W, Liang X, Nan K, Yao Y, Tian T. MicroRNA-155-5p promoteshepatocellular carcinoma progression by suppressing PTEN through the PI3K/Akt pathway. Cancer Sci2017;108(4):620-631.

35. Li DP, Fan J, Wu YJ, Xie YF, Zha JM, Zhou XM. MiR-155 up-regulated by TGF-beta promotesepithelial-mesenchymal transition, invasion and metastasis of human hepatocellular carcinoma cells in vitro.Am J Transl Res 2017;9(6):2956-2965.

36. Tokushige K, Hyogo H, Nakajima T, Ono M, Kawaguchi T, Honda K, Eguchi Y, Nozaki Y, KawanakaM, Tanaka S, Imajo K, Sumida Y, Kamada Y, Fujii H, Suzuki Y, Kogiso T, Karino Y, Munekage K,Kuromatsu R, Oeda S, Yanase M, Mori K, Ogawa Y, Seko Y, Takehara T, Itoh Y, Nakajima A, KanemasaK, Nishino K, Masaki N, Takahashi H, Seike M, Torimura T, Saibara T, Toyota J, Chayama K, HashimotoE. Hepatocellular carcinoma in Japanese patients with nonalcoholic fatty liver disease and alcoholic liverdisease: multicenter survey. J Gastroenterol 2016;51(6):586-596.

Figure Legends

Fig. 1 Illustration of the framework for mining regulatory pathways and key regulators by constructingHCV-induced HCC networks.

Fig. 2 A, B and C. The HCC, HCC-HCV and HCV-related TF-miRNA regulatory network GO and KEGGfunctional enrichment analysis. The numbers represent gene counts for each pathway/GO term within thenetwork. The Q-value represents the Bonferroni-corrected p-value in gene enrichment analysis. The richfactor represents the ratio of the number of genes in the subnetworks to the total number of genes in thepathways/GO terms. (A) Disease-related background network. (B) HCC-related TF-miRNA regulatorysubnetwork. (C) HCC-HCV-related TF-miRNA regulatory subnetwork. (D) HCV-related TF-miRNA reg-ulatory subnetwork.

Fig. 3 The key TF-miRNA regulatory network. The orange nodes represent target genes, the blue nodesrepresent TFs, and the green nodes represent miRNAs. Different border colors represent the gene source ofdifferent subnetworks.

Fig. 4 EZH2 gene differential expression significantly correlated with the overall survival of HCC patients.

Table 1 The KP score of predicted pathways from the three disease-associated subnetworks.

Type Pathway

HCC NR4A1-E2F1-hsa-miR-93-5p-EZH2-hsa-miR-155-5p-E2F2-hsa-miR-92a-3p-CDK1HCC NR4A1-E2F1-hsa-miR-93-5p-EZH2-hsa-miR-155-5p-E2F2-hsa-miR-92a-3p-HIST1H2AMHCC hsa-miR-26b-5p-FOXM1-hsa-miR-200b-3p-EZH2-hsa-miR-155-5p-CDKN2A-hsa-miR-141-3p-CDC25CHCC STAT3-FOXM1-hsa-miR-200a-3p-EZH2-hsa-miR-155-5p-PLK1-hsa-miR-141-3p-CDC25CHCC STAT3-FOXM1-hsa-miR-200a-3p-EZH2-hsa-miR-155-5p-PLK1-hsa-miR-141-3p-CDC25CHCC-HCV MYC-FOXM1-hsa-miR-200b-3p-MYB-hsa-miR-155-5p-CHD8HCC-HCV MYC-FOXM1-hsa-miR-200a-3p-MYB-hsa-miR-155-5p-RAP1BHCC-HCV MYC-FOXM1-hsa-miR-200b-3p-MYB-hsa-miR-155-5p-RAP1BHCC-HCV MYC-FOXM1-hsa-miR-200a-3p-MYB-hsa-miR-155-5p-SLC39A14HCV hsa-miR-155-5p-CBFB-APCHCV hsa-miR-155-5p-CBFB-hsa-miR-221-3p-BBC3HCV hsa-miR-155-5p-CBFB-hsa-miR-221-3p-GJA1HCV hsa-miR-124-3p-NR4A1-E2F1HCV hsa-miR-93-5p-NR4A1-E2F1

7

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

Figure 1

8

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y.

Figure 2

9

Pos

ted

onA

uth

orea

6M

ar20

20—

CC

BY

4.0

—htt

ps:

//doi

.org

/10.

2254

1/au

.158

3538

63.3

3858

930

—T

his

apre

pri

nt

and

has

not

bee

np

eer

revie

wed

.D

ata

may

be

pre

lim

inar

y. Figure 3

Figure 4

10